October 10, 2012 | 2
My high school biology teacher once told me that nothing was binary in biology except for alive and dead, and pregnant and not pregnant. Any other variation, he said, existed along a continuum. Whether or not the claim is technically accurate, it serves to illustrate an important feature of biological life. That is, very little in the biological world falls neatly into categories. A new finding, published today in PLoS ONE by Gustavo Arriaga, Eric P. Zhou, and Erich D. Jarvis from Duke University adds to the list of phenomena that scientists once thought were categorical but may, in fact, not be.
The consensus among researchers was that, in general, animals divide neatly into two categories: singers and non-singers. The singers include songbirds, parrots, hummingbirds, humans, dolphins, whales, bats, elephants, sea lions and seals. What these species all have in common – and what distinguishes them from the non-singers of the animal world – is that they are vocal learners. That is, these species can change the composition of their sounds that emanate from the larynx (for mammals) or syrinx (for birds), both in terms of the acoustic qualities such as pitch, and in terms of syntax (the particular ordering of the parts of the song). It is perhaps not surprising that songbirds and parrots have been extremely useful as models for understanding human speech and language acquisition. When other animals, such as monkeys or non-human apes, produce vocalizations, they are always innate, usually reflexive, and never learned.
But is the vocal learner/non-learner dichotomy truly reflective of biological reality? Maybe not. It turns out that mice make things more complicated.
Only in the last hundred years or so have researchers known that mice vocalize as part of their mating process. The reason it eluded scientists for so long is that their vocalizations can’t be heard by human ears. But then, in 2005, Holy and Guo argued in a paper in PLoS Biology that the ultrasonic vocalizations produced by mice ought to be thought of as songs rather than calls.
Lots of species produce calls, and those calls serve different purposes. Some are primarily used for mating, others for indicating the presence of food, and still others to notify group members of predators. While some calls may indeed be thought of as musical, scientists tend to distinguish between “calls” and “songs.” Unlike calls, which are built of single syllables (sometimes repeated), songs include multiple syllables that are constructed in a specific (non-random) order, often with repeated phrases. Calls tend to be identical across multiple individuals of a given species, while songs tend to differ from singer to singer.
The binary distinction between singers and non-singers might not be as convincing if it were based solely on observable behavior, but it turns out that the dichtomy is reflected in neurobiology. There are special neural circuits in both humans and singing birds that are uniquely associated with vocal learning.
Mice had always been situated firmly in the “vocal non-learning” group, but if Holy and Guo are right in referring to mouse vocalizations as songs, Arriaga and his colleagues reasoned, then they might show the same neurobiological signature as birds and humans. One of the hallmark neurobiological features of song learners is a circuit that starts in the motor cortex on the top of the brain which projects directly to the part of the brainstem that controls the vocal organ. These circuits have never been seen in any other non-singing species, according to Arriaga, “despite over fifty years of effort searching for them, particularly in vocal non-learning birds and non-human primates.”
The researchers discovered that mice do have a brain circuit that starts in the primary motor cortex, projects directly to the part of the brainstem responsible for controlling the larynx, and importantly, is active when male mice sing. The difference, when compared with birds and humans, is that the circuit is weaker, more sparse. It’s there, it’s just not as strong.
When this pathway is disrupted in singing birds or humans, they become unable to produce vocalizations that had been learned (songs), but are still able to produce their innate vocalizations (calls). So Arriaga wanted to see what would happen if he chemically disabled those circuits in some mice. While the impaired mice were still able to sing their songs, they didn’t sound quite right. Both the pitch and the frequency of their vocalizations had been affected.
Another difference between singers and non-singers is that when a singer becomes deaf, their songs (and speech, for humans) deteriorate, but the same is not true for those whose vocalizations are innate. Deafened monkeys and non-singing birds do not show any reductions in their abilities to produce vocalizations. What this means is that vocal learners require auditory feedback in order to maintain their songs. And like humans and singing birds, mice that were deafened showed a gradual deterioration of the quality of their songs, a process that continued over several months. And congenitally deaf mice who had never been hearing were terrible singers; their songs “sounded like squawks and screams rather than whistles.” Together, the experiments with artificially and genetically deaf mice show that they need experience to learn their songs in the first place, and that they require continual experience to maintain them.
The researchers also noticed, quite by accident, that the different strains of mice had songs with predictably different pitch. In a way, the different strains of mice had their own accents. They had already shown that mice require auditory feedback to maintain the acoustic qualities of their songs, but now they wondered if they can explicitly modify those features. In other words, were mice capable of vocal improvisation? Could they adjust their accents? To see if they could improvise, the researchers put male mice of different strains into the same housing area. Before the experiment, the songs of the two strains differed by at least six thousand hertz, or as much as nine thousand hertz. After eight weeks, the mice had adjusted the pitch of their songs, reducing the pitch difference by at least half. Those who started with higher pitch lowered it over time to match their new neighbors, while those with lower pitch likewise raised the pitch of their songs. Several of the pairs narrowed the pitch difference to less than five hundred hertz.
Altogether, the researchers demonstrated that mice have each of the features traditionally associated with vocal-learning song-producing species, but also that for each of those features they are not quite as sophisticated or advanced as humans or song-learning birds. What is striking, however, is that those features are not completely absent, as was previously assumed. The strict dichotomy between singers and non-singers, or more accurately, between vocal learners and non-learners, may not be as strict as once was thought.
Instead, the researchers propose that this phenomenon may exist “along a continuum, with vocal mimics and some supposed vocal non-learning species at either extreme.” Mice may occupy an intermediate space between species for whom all vocalizations are innate, and those which may be more similar to vocal learners such as humans, birds, whales, and others. The extent to which mice do learn their vocalizations in a manner similar to humans, however, means that scientists who study disorders that feature impairments of speech or language, such as autism, Parkinson’s, Tourette’s, and others, may find them even more useful as a model species. Given the ubiquity of the mouse in genetic laboratories around the world, they may become useful more generally, as scientists continue to uncover the evolutionary origins of language.
Arriaga G, Zhou EP, Jarvis ED (2012) Of Mice, Birds, and Men: The Mouse Ultrasonic Song System Has Some Features Similar to Humans and Song-Learning Birds. PLoS ONE 7(10): e46610. doi:10.1371/journal.pone.0046610
Header image via Wikimedia Commons/George Shuklin.